Qualcomm's M1 Rival to Be Available in PCs by Late 2023

[BanjoDudeAhoy]

BUt MS has to support all of those as well because there are significantly more applications that were built and developed for windows over the years that buisnesses rely on. Signficantly more buisnesses use Windows and or windows Server to handle loads internally.

So any type of "rosetta" for windows would need to be more complex.

The reason Rosetta Works well with MacOS is that over time Mac simply just was like. welp.... we wont let 32 bit apps run now, we wont let this run now etc ertc. Then with monty and M1 They simplified what rosetta 2 had to do.

Windows on the other hand supports lots of legacy applications. And a lot of those applications are not build on mac and no other alternative exists.


Windows already many years back had a somewhat decent arm version of windows on the surface 2 RT. But that is only if the application was built for the ARM version of it. Decently power efficient. But there was not a market to keep developing the arm platform back then. The newer Surface with the newer ARM chips have a built in translator that runs dog slow. The sheer process of trying to make all legacy applications compatible is going to be a monumental undertaking compared to rosetta 2.

This is the point I was trying to make earlier. You worded it better.
I think legacy support on Windows is a great thing (just thinking of all the old games I can run without having to rely on re-releases or remasters) but I also think it’s holding it back, preventing it from “just” making the sort of jump Apple did with Apple Silicon.

 

[mannyvel]

Qualcomm/ARM can kill off all the other embedded chips, like the PowerPC, MIPS, and the various "low power" x86 stuff. And it'll bring more power to the smaller embedded systems. It's basically a win all around.

It's weird that Qualcomm doesn't have the expertise to do this in-house ie: they had to buy a company of ex-Apple employees to make this happen. It makes you wonder how crappy the development for the rest of their stuff is.

 

[planteater]

You think Qualcomm isn't working closely with Microsoft on this?

Also, remember: Qualcomm's products are mobile-first. These will probably go into laptops and 2-in-1 devices, exclusively. Don't think these will go into desktops with their 100W+ power supplies and large thermal envelopes, like the M1 Ultra.

So, M1 and Nuvia would be a direct comparison. Let's see how they do.

Why wouldn’t they go into desktop/server class machine? Enterprises running Windows Server stack want energy savings as well. It will absolutely be in those machines at some point in the roadmap.

 

[GMShadow]

Ok, that's fine if pricing is well below apple. Android making real change for the many.

Yeah, I’m sure Qualcomm really cares about $50 devices for India.

 

[macsplusmacs]

Android making real change for the many.

"real change" for fewer and fewer every quarter.

Apple Was the Only Top Smartphone Vendor to Increase Shipments Year-Over-Year Last Quarter

Apple was the only top smartphone maker that saw a year-over-year increase in shipments in the last quarter, while Samsung, Oppo, Xiaomi, and others...

www.macrumors.com

 

[deathcab]

in 2023-2024, Apple will be far ahead. Qualcomm CEO needs to know Apple is already working on M2 and M3 Apple Silicon Chip, ?

You are 100% correct. I’m sure that’s Qualcomm, one of Apple’s largest suppliers and competitors who bought a company founded by former Apple chip engineers, has no idea Apple is working on their next generation of chips. /s

 

[OnawaAfrica]

Is that really in Microsoft's interest to do so?

thats microsofts own problem if they dont wana do that. but its not blocked by apple

 

[OnawaAfrica]

The other side of this coin is “you can’t run old apps that might still be perfectly useful and your business might depend on.”

---

Edit: Just Seen Mojave is not supported.

 

[DrV]

I’m sure China will be at it as well, probably named the Xi1, Xi1 Pro, Xi1 Max, Xi1 Ultra and Xi1 Super Mega Better Than Apple

China has a major obstacle here. Their problem is called ASML.

ASML is a Dutch company manufacturing semiconductor lithography tools. It is the only company supplying EUV tools needed for low-nanometer chip manufacturing. It does not sell its tools to China.

This keeps Chinese chipmakers locked to the older DUV technology. It works well for most uses, but does not compete in complex low-power applications, such as large CPUs.

 

[Radeon85]

Apple will be on M3/M4 by then and far ahead, saying that you will have an M1 rival by 2023/2024 is nothing to brag about.

 

[jakey rolling]

you are very wrong. x86 / x64 are very different in design Structure and Architecture. RISC/CISC has nothing todo with Compatibility Layers, they are complete different instruction sets. The Problem with x86/64 is not the Legacy ISA, its the whole design it self. One issue is the Power Inefficiency and the huge amount of Heat Generated.

As someone who actually lived through the PowerPC days, this whole song and dance sounds eerily familiar.

I remember Apple crowing about the "inefficient" Pentiums of the time. The PowerPC architecture was supposed to be this massively power-efficient speed demon that "toasted" those poor Intel bunnies and let Apple users run desktops and laptops at speeds that x86 users could only dream of. All those benchmarks showing how the "Megahertz Myth" meant that those 233 MHz G3s were so much faster than 450MHz Pentium II's, and used much less power.

We know how that whole thing ended up. Nearly constant CPU production shortages, "Wind Tunnel" G4s, loud liquid-cooled G5s, no 3GHz machines, and no G5 laptops. And then an inevitable switch to the supposedly "inferior" x86 processor line.

 

[macsplusmacs]

As someone who actually lived through the PowerPC days, this whole song and dance sounds eerily familiar.

I remember Apple crowing about the "inefficient" Pentiums of the time. The PowerPC architecture was supposed to be this massively power-efficient speed demon that "toasted" those poor Intel bunnies and let Apple users run desktops and laptops at speeds that x86 users could only dream of. All those benchmarks showing how the "Megahertz Myth" meant that those 233 MHz G3s were so much faster than 450MHz Pentium II's, and used much less power.

We know how that whole thing ended up. Nearly constant CPU production shortages, "Wind Tunnel" G4s, loud liquid-cooled G5s, no 3GHz machines, and no G5 laptops. And then an inevitable switch to the supposedly "inferior" x86 processor line.

Good point, but wasn't Motorola the main company in charge during the PowerPC days?

This time around though it's all Apple for Apple Silicon. With a VERY deep CPU design bench, more brilliant people, more Fab operation options and more money than god.

 

[jakey rolling]

Good point, but wasn't Motorola the main company in charge during the PowerPC days?

This time around it's all Apple for Apple Silicon. With a VERY deep CPU design bench, more brilliant people, more Fab operation options and more money than god.

The PowerPC was originally the work of the "AIM Alliance": (A)pple, (I)BM, and (M)otorola. I recall IBM and Motorola both produced G3 and G4 processors at different times. IBM was the main company behind the G5, and I recall (at least based on its announcement) that IBM built a US-based fabricating plant specific for that processor.

At the same point in its existence, PowerPC (and other RISC architectures, like SPARC and Alpha) also showed incredible promise, and the common thought during that time was that these various architectures were going to supplant x86 in the very near future. Even Intel was trying to replace x86 with a successor when they introduced Itanium. Apple was the only manufacturer to turn out a consumer computer running a "modern" RISC CPU.

My only point with this was just that what we're seeing and hearing about ARM supplanting x86 and Apple taking the lead with an incredible CPU design that is leaving "old, inefficient" x86 architecture in the dust echos what was being said ~25 years ago.

 

[macsplusmacs]

The PowerPC was originally the work of the "AIM Alliance": (A)pple, (I)BM, and (M)otorola. I recall IBM and Motorola both produced G3 and G4 processors at different times. IBM was the main company behind the G5, and I recall (at least based on its announcement) that IBM built a US-based fabricating plant specific for that processor.

Oh yeah. forgot about AIM.

 

[chucker23n1]

As someone who actually lived through the PowerPC days, this whole song and dance sounds eerily familiar.

I remember Apple crowing about the "inefficient" Pentiums of the time. The PowerPC architecture was supposed to be this massively power-efficient speed demon that "toasted" those poor Intel bunnies and let Apple users run desktops and laptops at speeds that x86 users could only dream of. All those benchmarks showing how the "Megahertz Myth" meant that those 233 MHz G3s were so much faster than 450MHz Pentium II's, and used much less power.

We know how that whole thing ended up. Nearly constant CPU production shortages, "Wind Tunnel" G4s, loud liquid-cooled G5s, no 3GHz machines, and no G5 laptops. And then an inevitable switch to the supposedly "inferior" x86 processor line.

Yes and no. On the one hand, I've always found the distinction and benefits of "RISC" to be overstated.

On the other hand, it's tricky to use PowerPC as an example. By the late 1990s, several things had happened to AIM:

• Apple was recovering from almost getting acquired or going bankrupt. It was not yet clear if the Mac was going to see much of a renaissance.
• IBM had, at this point, had a terrible time with PCs. Their x86 desktops weren't going anywhere, neither was OS/2, and their PowerPC workstations were mostly stillborn. Their laptop business was successful enough that they bothered selling it to Lenovo in 2005. To them, PowerPC was ultimately mostly useful as a vehicle to advance POWER, for server use. PowerPC 970 ("G5") was a bit of an exception to that, but clearly didn't go where either Apple or IBM wanted it to. Then there's also the various game consoles: the 970-based Xbox 360, various 750 ("G3")-based variants Nintendo used in the GameCube, Wii, Wii U, and oddest of all, the Cell in the Playstation 3.
• Motorola was all over the place, including participating in Apple's Mac clone program. (There's a case to be made that, if Apple hadn't canceled that, the PowerPC relationship might've panned out quite differently?) They also made CodeWarrior, the preëminent IDE for PowerPC development. But then their interest shifted towards only really wanting to use PowerPC for embedded stuff like networking hardware, and by 2004, they spun it out altogether.

Given all that, the question becomes: who's rooting for PowerPC chips with good desktop performance? Neither IBM nor Motorola are selling desktops, so not them. Who's rooting for power-efficient PowerPC chips for laptops? IBM has at this point given up and is only doing x86 laptops, so not them. Certainly not Motorola either.

All of that is quite different from the early 1990s, when Motorola was a lot more involved with PC CPUs (the 68k was also in NeXT, Amiga, and so many others), and when IBM not only had multiple collaborations with Apple, such as Taligent, but also thought there was a path forward where they could leverage PowerPC against the Wintel alliance. That never happened.

So Apple had the G3 and G4 which still had OK efficiency (but increasingly poor performance) for their laptops, and the G5 which had OK performance (but poor efficiency) for their desktops.

None of that is a factor today. There's no confusion about the relationship: Apple licenses the ISA from ARM (it's unclear if they'll even bother for ARMv9), then designs custom cores, then has TSMC manufacture them. The only pressure they need to put is for TSMC to keep advancing their manufacturing, and so far, that's been working out.

 

[Nuno Lopes]

Why don’t these companies just stop announcing and actually deliver? Don’t they understand that this is the same as saying … “I have nothing!”


A radical concept maybe?

 

[williedigital]

The PowerPC was originally the work of the "AIM Alliance": (A)pple, (I)BM, and (M)otorola. I recall IBM and Motorola both produced G3 and G4 processors at different times. IBM was the main company behind the G5, and I recall (at least based on its announcement) that IBM built a US-based fabricating plant specific for that processor.

At the same point in its existence, PowerPC (and other RISC architectures, like SPARC and Alpha) also showed incredible promise, and the common thought during that time was that these various architectures were going to supplant x86 in the very near future. Even Intel was trying to replace x86 with a successor when they introduced Itanium. Apple was the only manufacturer to turn out a consumer computer running a "modern" RISC CPU.

My only point with this was just that what we're seeing and hearing about ARM supplanting x86 and Apple taking the lead with an incredible CPU design that is leaving "old, inefficient" x86 architecture in the dust echos what was being said ~25 years ago.

At the end of the day, it doesn't matter much now. Even craphole PCs are virtually silent and fast enough for cloud-based apps that rule now. Until there are meaningful new applications for the consumer space, it's all unimportant.

 

[chucker23n1]

Why don’t these companies just stop announcing and actually deliver? Don’t they understand that this is the same as saying … “I have nothing!”

In this case, because it's hard to brush off analysts without risking shareholder value.

Qualcomm has published its latest set of financials (concerning Second Quarter Fiscal 2022 Results), followed up with a lengthy earnings call, with answers provided by President and CEO, Christian Amon, among others. During the call Amon provided an update on the integration of the Nuvia team, and progress in their goal of developing a big-leap-forward for Arm processors. Amon commented upon the timescale in which we should expect to see Nuvia powered Windows laptops on retail shelves, indicating things are currently on track for late 2023.

In answer to a question from an IT industry analyst regarding the laptop market and Qualcomm silicon, Amon said that he was happy with the commercial and enterprise designs of Windows 11 systems based around the newest Snapdragon 8CX Generation 3. Then he brought up the Nuvia designed processor, which he claimed will be "going after the performance tier… development is on track, and we expect to have that in late 2023."

Apple can do this because they've been disciplined about it for a quarter century, and because it's what analysts (mostly) expect of them by now.

 

[williedigital]

Yeah, I’m sure Qualcomm really cares about $50 devices for India.

Android has majority market share in every market in the developed world, including the USA.

 

[chucker23n1]

Android has majority market share in every market in the developed world, including the USA.

…and?

 

[OnawaAfrica]

As someone who actually lived through the PowerPC days, this whole song and dance sounds eerily familiar.

I remember Apple crowing about the "inefficient" Pentiums of the time. The PowerPC architecture was supposed to be this massively power-efficient speed demon that "toasted" those poor Intel bunnies and let Apple users run desktops and laptops at speeds that x86 users could only dream of. All those benchmarks showing how the "Megahertz Myth" meant that those 233 MHz G3s were so much faster than 450MHz Pentium II's, and used much less power.

We know how that whole thing ended up. Nearly constant CPU production shortages, "Wind Tunnel" G4s, loud liquid-cooled G5s, no 3GHz machines, and no G5 laptops. And then an inevitable switch to the supposedly "inferior" x86 processor line.

ya true i used to have my self a PowerMac 8600/300 with a G3 CPU Upgrade. those old days of Mac OS 9.2.1

 

[visualseed]

Throwing money and resources at the problem won't give you a better CPU so this is the wrong metric for forecasting success. This is why AMD produced a significantly better x86 CPU with 1/10th the resources of Intel. Samsung has unlimited resources but have not produced a leading mobile application processor yet. The team that created the original CPU that Nvidia is using for data centers had 15 people. Google has unlimited resources and even a strong brand but their chip team is a mess. There are a lot of steps in the CPU creation process that require quality over quantity. Quality of talent is everything in CPU design and that's not something you can scale easily.

One thing that people should take note of is that Qualcomm still trails Apple in mobile CPU performance and they have been at it for decades. If Qualcomm had a chance to beat Apple in desktop CPU performance in the near future, I think we would already see evidence of that in their mobile processors.

 

[FriendlyMackle]

Apple friends-don’t sleep on what Qualcomm is doing. They are a major company that has lots of resources and also the resources of Microsoft. Microsoft is desperate for a platform to run Windows for arm on. They are working with Qualcomm on its platform and also developing a platform in-house.

They fully understand that x86 is a old,tired ,outdated architecture.Apple made this jump and they got there first because they could leverage the work that they had done for the iPhone and iPad. Now the pc world is desperately looking to catch up.
Microsoft is throwing money at this in a very big way.
if you are only interested in the Mac, this will still have a effect on you as Apple has to respond to what happens on the PC side of the computer business.

But please don’t just write this off and say that Apple is so far ahead that no one can catch up. Microsoft can and will catch up. They will try very hard to surpass what Apple has done. It’s still a open question if they can pass Apple, but I have been working with them for around 30 years and I will say that I won’t ever bet against them.

I agree with everything you’ve said—except that I don’t see Microsoft surpassing Apple in terms of the quality of their software/processors/hardware (to whatever degree they involve themselves in direct design/manufacturing) despite their massive resources. I do think that Microsoft will reach parity…whether that will translate into a real-world lead is more questionable.

As others have said, Microsoft is really hampered by backwards-compatibility with products spanning decades. Perhaps they will create a new walled-off ‘next-generation’ line of products. Perhaps that will be more successful than their earlier attempt.

I don’t know how Microsoft manages teams internally, but for all that can be said for Apple’s sometimes (seeming) myopia focusing on only one or two products at a time, Apple’s approach has yielded industry-changing results repeatedly over the past three decades…from phones to laptops to small form-factor desktops, and all-in-ones (and ignoring their earlier strides in introducing home computers to the masses in the 1980s!).

 

[FriendlyMackle]

China has a major obstacle here. Their problem is called ASML.

ASML is a Dutch company manufacturing semiconductor lithography tools. It is the only company supplying EUV tools needed for low-nanometer chip manufacturing. It does not sell its tools to China.

This keeps Chinese chipmakers locked to the older DUV technology. It works well for most uses, but does not compete in complex low-power applications, such as large CPUs.

I am sure that China is working on stealing those lithography tools. We know that industrial espionage is a common threat—now consider that in the case of Chinese firms, this type of espionage is backed by the resources of the state. I’m sure that the Chinese government consider this a massive security failing of dependence upon a foreign manufacturer/designer for next-generation technology implementation. I hope that ASML is able to protect their processes and IP.

 

[visualseed]

The PowerPC was originally the work of the "AIM Alliance": (A)pple, (I)BM, and (M)otorola. I recall IBM and Motorola both produced G3 and G4 processors at different times. IBM was the main company behind the G5, and I recall (at least based on its announcement) that IBM built a US-based fabricating plant specific for that processor.

At the same point in its existence, PowerPC (and other RISC architectures, like SPARC and Alpha) also showed incredible promise, and the common thought during that time was that these various architectures were going to supplant x86 in the very near future. Even Intel was trying to replace x86 with a successor when they introduced Itanium. Apple was the only manufacturer to turn out a consumer computer running a "modern" RISC CPU.

My only point with this was just that what we're seeing and hearing about ARM supplanting x86 and Apple taking the lead with an incredible CPU design that is leaving "old, inefficient" x86 architecture in the dust echos what was being said ~25 years ago.

We are almost 15 years into Apple using ARM on phones and more than a decade of them using their own design exclusively. The PowerPC chips were relegated more by IBM not seeing it a viable business for a single customer than by performance or thermal issues (which did exist too). In fact, they eventually rolled the PPC instruction set back into their multicore POWER chips. Which is now a beast of a mainframe processor called the Power10 and in performance class all to itself. So the technology still has legs despite not being in Macs.

While G5s were not the ice-cold miracles Apple originally touted PPC would be but nor were the Xeons that replaced them. I'm not going to wager the ARM will be the death knell for X86/64, but it's probably not going be anything less than significant competition and a viable alternative performance-wise and the market leader for power efficiency for a very long time.